Central Statistical Bureau of Latvia (CSB)

Business Statistical Department, Industrial and Construction Statistics Section.

Lacplesa Street 1, Riga, LV - 1301 

The Construction Production Index (IPC) is an important indicator for monitoring and evaluating the development of the construction sector and is widely used in the analysis of business cycles. The production index for construction measures changes over time in the price-adjusted output of construction (Section F of NACE Rev. 2). It provides an approximation measure of the volume trend in value added over a given reference period.

The index is obtained for

• Total construction works
• Construction of Buildings
• Construction of Civil engineering

and presented in the following forms:

• Unadjusted
• Calendar adjusted
• Seasonally adjusted

The IPC is also used to provide data for national accounts for the calculation of GDP.

NACE Rev. 2. classification.

Activities covered:

IPC covers all economically active enterprises the main or secondary activity of which in compliance with the NACE Rev. 2 is Construction (Section F, classes 41.20–43.99).

Size classes covered:

Units with at least 20 employees are surveyed by a full survey. Other commercial companies engaged in construction are surveyed based on stratified simple random sampling.

Small enterprise (with less than 5 employees) data are estimated with administrative data (VAT data from State Revenue Service).

Transmitted data to Eurostat:

Eurostat uses the CC classification (Classification of Types of Construction), which breaks down in Buildings and Civil engineering construction.

Considering the use of the NACE classification, the data to Eurostat are transmitted in the following way:

- NACE Rev.2 section F  “Construction”;

- F_CC1 “Buildings” (NACE Rev.2 division 41);

- F_CC2 “Civil engineering works” (NACE Rev.2 division 42).

Construction production indices characterize changes of construction production during the reference period in comparison with the base period. Construction output consists of the amount of own-account construction works completed by the construction enterprises.

Value of construction work done during the reference period is aggregated into NACE classes and deflated by corresponding construction cost indices separately. If the production data is not available turnover data is used instead. VAT is excluded.

Calculation of IPC is based on the chain index method; within the framework of it the average quarterly volume of construction work done (at constant prices) previous year is used as a calculation basis and the value added of construction enterprises produced two years ago is used as weights.

Reporting unit - enterprise

Observation unit - KAU

The Statistical Business Register based list of active economic units which are active in the field of construction. There are about 9000 enterprises in the target population.

Reference area are construction works done in Latvia and abroad.

Time series for this indicator according to NACE Rev. 2 is available from 2000.

Base year: 2015

Index number

Quarter

Official Statistics Law is the legal base of all statistics (https://likumi.lv/ta/id/274749-statistikas-likums)

Rules of Cabinet of Minister of the Republic of Latvia No. 691 of 24 November 2020 concerning to National Statistical Information program for the 2021-2023.

Regulation (EU) 2019/2152 of the European Parliament and of the Council of 27 November 2019 on European business statistics, repealing 10 legal acts in the field of business statistics

The former legal basis for the STS indicators is the Council Regulation No 1165/98 of 19 May 1998 concerning short-term statistics and subsequent amending regulations.

Commission Implementing Regulation (EU) 2020/1197 of 30 July 2020 laying down technical specifications and arrangements pursuant to Regulation (EU) 2019/2152 of the European Parliament and of the Council on European business statistics repealing 10 legal acts in the field of business statistics 

EUROSTAT

Confidentiality of the information provided by respondents is protected by the Chapter III Section 17 of the Official Statistics Law stipulating rights and obligations of the Central Statistical Bureau and other state authorities producing official statistics.

Section 17. Data Processing and Statistical Confidentiality

(1) The statistical institution shall use the data that are obtained in accordance with Sections 14, 15, and 16 of this Law for the production of official statistics, with the exception of cases laid down in Sections 25 and 26 of this Law.

(2) The data may be modified, supplemented or combined with other data available to the statistical institution, as well as re-used for the production of other official statistics.

(3) In using data, the statistical institution has the right to establish and maintain statistical registers for the production of official statistics.

(4) The statistical institution shall determine the data storage duration, by taking into account the needs for production of official statistics.

(5) Data of a natural person obtained for the production of official statistics shall be anonymised immediately after collection, testing and linking thereof, except the case where personal data are still required for production of official statistics. The statistical institution shall ensure that the identifying data are stored safely and separately from other data.

(6) The statistical institution shall take all the necessary measures to prevent unauthorised access to data, data modification or dissemination, accidental or unauthorised destruction.

(7) Employees of the statistical institution shall not disclose data or any other restricted access information coming to their knowledge in the performance of their service or work duties. This shall also apply to persons who are temporarily involved in the production of official statistics or have terminated employment or service relations.

Statistical data shall be considered confidential if they directly or indirectly allow for identification of the private individuals or State authorities regarding which personal statistical data have been provided. Confidential data are treated by suppression. Data are confidential if there are 3 or less enterprises or one enterprise accounts for more than 80% or two of them more than 90%. Also, the data of individual enterprises or establishment are confidential.

The final data on index of construction are published on the 40^nd day after the end of the reference period according to Release calendar. Dissemination calendar is available on the Official statistics portal: 

https://stat.gov.lv/en/calendar?Dates=%22Next+8+days%22

On construction and building permits granted:                              

Reference period:               Release date:

1st quarter  2022                   10-May-2022

2nd quarter 2022                   09-Aug-2022

3rd quarter 2022                    09-Nov-2022

4th quarter 2022                    09-Feb-2023

Dissemination calendar is available on the Official statistics portal: https://stat.gov.lv/en/calendar?Dates=%22Next+8+days%22

The data are released simultaneously to all interested parties by issuing the press release "On construction and building permits granted" in xxx (quarter)" both to the media and to the main government officials.

Data are transmitted to Eurostat on the same day that the data are published in the national publication, 40 days after the reference period.

Quarterly.

Statistical information is published in accordance with an approved release calendar. It's available on the Official statistics portal: https://stat.gov.lv/en/calendar?Dates=%22Next+8+days%22

The data are released simultaneously to all interested parties by issuing a press release at 13.00 o' clock of the local time. At the same time the data are also sent to subscribers by e-mail. The data set is subsequently posted on the Official statistics portal. The press release and the indicators are also published on the Internet.
The CSB publishes quarterly and annual data of the construction production index on the Internet website: https://stat.gov.lv/en/statistics-themes/business-sectors/construction

The Construction production index is published in "Statistical Yearbook of Latvia" - annually. 

Information on construction production indices is available on the Official statistics portal homepage database under section Construction and real estate.

Comparable time series are available since 2000 (with the base year 2015).

The Official statistics portal internet database https://data.stat.gov.lv/pxweb/en/OSP_PUB/START__NOZ__BU__BUP/?tablelist=true

Volume indices and changes in volume of construction production by economic activity:      Link to database table.

• 2015=100, seasonally adjusted data
• 2015=100, calendar adjusted data
• 2015=100, non-adjusted data
• % over the previous year, calendar adjusted data
• % changes compared of previous period, seasonally adjusted data

Construction output at current prices by economic activity, thsd euro:     Link to database table.

• NACE Section, Division, Groups and some classes.

Construction of buildings output at current prices by building type, thsd euro (sub-table of NACE 4120 from the preceding table):     Link to database table.

• Residential buildings (Total and New construction) by subtype of buildings
• Non-residential buildings (Total and New construction) by subtype of buildings

Not available.

Data are sent to Eurostat both to be used in European aggregates as well as to be released as national data. Data are available on the Eurostat database in section Industry, trade and services/Short - term business statistics, link: http://ec.europa.eu/eurostat/data/database

The metadata section of the Official statistics portal includes a description of the methodology: https://stat.gov.lv/en/metadata/8073-volume-index-production-construction

From 2021, standardized quality indicators are placed in the detailed metadata section of the Official statistics portal : https://stat.gov.lv/en/metadata/8073-volume-index-production-construction/sims2

The quality of statistics is assessed according to five quality criteria: relevance, accuracy, timeliness and punctuality, accessibility and clarity, coherence and comparability.
The quality indicators are calculated taking into account Eurostat defined methodology for the calculation of the Quality indicators and documented into Project's Documentation System.

Every quarter on the Project's Documentation System, fields of relevance, accuracy and reliability, timeliness and punctuality, comparability, coherence are completed.

The regulations and recommendations governing the calculation of the construction production index specify the essentiality, methodology, level of detail, and timely reporting of the index. Any attempt to introduce supplementary indicators or fields in the reporting form can impose a heavier workload on the respondents and potentially compromise the quality of the data. This is primarily because the information necessary to complete the report may not be readily available to the respondents, leading to delays in reporting, which can, in some cases, have significant consequences.

The main users are:

• Eurostat, OECD, UN, etc.
• Bank of Latvia
• Ministry of Economy
• Ministry of Finance
• Media
• Enterprises
• Public
• Other CSB departments (internal)

CSB's operational mission is to provide users of statistical information with independent, high-quality official statistics for decision-making, research and discussion.

Anyone can submit an opinion on the quality of the data or make any other suggestion via e-mail address: pasts@csp.gov.lv

Statistics on construction production indices are available on a quarterly basis in accordance with the requirements of Regulation (EU) 2019/2152 of the European Parliament and of the Council of 27 November 2019 on European business statistics, repealing 10 legal acts in the field of business statistics. The former legal basis for the STS indicators is the Council Regulation No 1165/98 of 19 May 1998 concerning short-term statistics and subsequent amending regulations.

As an indicator data completeness rate is defined as the the number of data cells (entities to be specified by the Eurostat domain manager) provided to the number of data cells required by Eurostat or relevant. The ratio is computed for a chosen dataset and a given period.The indicator shows to what extent statistics are available compared to what should be available.

For IPC, the value of this indicator matches the target value, which means that 100% of the required or relevant data is available. More detailed information than is required from eurostat is also available on the national statistics portal.

Statistical data is always analyzed in detail and validated using automatic validation procedures that are updated quarterly based on the shortcomings identified. During the data collection process, all individual data are examined in detail to guarantee the quality and reliability of the data obtained, as well as compliance with methodological guidelines. Any errors, omissions or abnormal values ​​identified are corrected. At the time when data are provided to Eurostat and published nationally the response rate is the same. Telephone contacts are used to follow up missing responses after the 15th day of the quarter. Imputation is used to compensate the non-response; this is made by means of previous replies and/or the replies of other businesses. Sample size made so that the coefficient of variation doesn't exceed 1.83%. The coefficient of variation is calculated for total construction production (value of own account construction work). 

Accuracy is achieved by eliminating non-sampling errors as much as possible. Although measures are being taken to update the sample, the main sources of error are the level of non-response and overcoverage. Given the deadlines set for the collection and processing of the report, the non-response rate to the rapid assessment tends to be high. However, the report forms are still active after the submission deadline. This provides an opportunity to submit or update the data, which reduces the level of non-response.

IPC measurement error occurs in cases where the respondent misunderstands the required information or submits inaccurate information that does not comply with the reporting methodology. Such errors are corrected through automatic validations, while other identified errors are corrected by contacting the respondent. When filling in the e-report, the respondent has the opportunity to get acquainted with detailed instructions for filling in the report, which are regularly updated based on the identified shortcomings in filling in the report. In case the respondent refuses to make the correction, an imputation is made based on other available information, including information found in the administrative data.

When resolving cases of non-response by units, automatic reminders are sent to respondents about delays in submitting data. If the respondent does not provide data, an attempt is made to contact him by telephone. In cases where the respondent categorically refuses to submit data, the data is imputed.

Processing errors in the preparation of construction indicators are rare and are identified in a timely manner. Processing errors can occur in the process of data aggregation if the appropriate field attributes are not set for individual units, fulfilling the necessary conditions for algorithm operation. A small part of the data is imputed and the imputations are calculated based on the time series of the respondent's historical data, administrative data, data submitted in other reports or data of similar companies. The calculated imputation values ​​are evaluated by an expert and a decision is made to correct them if necessary. Thus, errors can also occur in the imputation coding process when assigning an inappropriate imputation code, as well as in entering the imputations themselves when there is a significant difference between the fast and final estimate values. Some processing errors may also occur during the processing of microdata for the sizing of a third part of the population, in cases where there are enterprises that have historically been included in construction but no longer carry out actual construction work. When identifying such companies, the administrative data are adjusted. In some cases, there are often respondents who adjust or refine the values ​​submitted to the final estimate, which creates a difference between the final estimate and the quick estimate. In cases where such actions have a significant impact on the value of the index, the respondent is contacted to find out the reason for the change.

The sampling error is usually mentioned as a part of the difference between a population value and an estimate thereof, derived from a random sample, which is because only a subset of the population is enumerated. Adjustments for non-response, outlier treatment and other uncertainty sources are included. 

The coefficient of variation (CV) is defined as the standard error divided by the expected value of the estimator. It is the standard error in relative (percentage) terms. It is the most suitable sampling error statistic for quantitative variables with large positive values, which are common in economic statistics. The smaller the CV, the standard error, the more accurate is the estimator. This means small sampling variability with regard to the estimate.

Coefficients of variation of variable, 2022:

Q1 1,38%

Q2 1,20%

Q3 0,92%

Q4 1,12%

Non-sampling errors are those errors in survey estimates which cannot be attributed to sampling fluctuations. Usually this type of error is divided into: coverage error, measurement error, non-response error and processing error.

Unit non-response rate - The ratio of the number of units with no information or not usable information (non-response, etc.) to the total number of in-scope (eligible) units. The ratio can be weighted or un-weighted. There can be many different reasons for reasons nonresponse as for example, refusal of the sample unit to participate, failure of the data collector to locate/identify the sample unit, failure to make contact with the sample unit; inability of the sample unit to participate (e.g. ill health, absence, etc); inability of the data collector and sample unit to communicate (e.g. language barriers); Accidental loss of the data/ questionnaire.

Un-weighted:          Weighted:

(2022)                      (2022)

Q1 14%                       22%

Q2 15%                       24%

Q3 15%                       26%

Q4 17%                       28%

Coverage error (or frame error) is due to divergences between the survey population and the target population.

Over-coverage rate - is defined for probability surveys and is the proportion of units accessible via the frame that do not belong to the target population (are out-of scope).

Un-weighted:          Weighted:

(2022)                      (2022)

Q1 1,2%                       2,9%

Q2 1,6%                       3,7%

Q3 2,0%                       4,0%

Q4 2,3%                       4,8%

The construction production index is published at 13:00 o' clock of local time on the 40^th day after the end of the reference quarter; if the 40th day is on a weekend or on a holiday then it is published on the next working day. The questionnaires are sent to the reporting unit at the end of the year for all quarters of the coming year, asking for a reply for each reference quarter till the 15th of the following quarter. Data collection is normally closed 7 days before the publication.

The data are published according to release calendar and sent to Eurostat at the same date when published press release in CSB website.

In 100% of cases data are published in line with the schedule.

For the whole Latvian territory, the same statistical concepts are applied.

Statistical information is comparable between EU countries.

The Construction production index is comparable since year 2000.

Differences between STS and SBS are explained as:

• Variable Construction production in STS is calculated taking into account only construction works done on own account, however in SBS construction activities also includes the work of subcontractors;
• STS sometimes do not cover enterprises with small number of persons employed but with large turnover (very often in activity “Development of building projects”), such enterprises frequently are detected and imputed only, when compiling SBS results;
• Different sample size, different unit of observation (STS-KAU, SBS- enterprise).
• Despite the fact that an enterprise may have turnover, this does not mean that works have actually been carried out, since companies often do not exclusively carry out certain NACE (F) related works. For example (author's supervision, supervision of the construction works, making projects, etc.) which accounts for a certain part of the company's turnover.

Nacional Acconunts use STS data for GDP calculation, mostly there are coherence between data.

Aggregates of IPC are not always coherent with their main sub-aggregates. Usually selection of seasonal adjustment model helps to achieve the coherence.

In Latvia, statistical questionnaires contains a field about the time necessary for respondents to fill in the questionnaire. According to the results, the average burden of respondents in 2022 was 0.39 hours per respondent. 

In line with the strategic directions of the European Statistics System and latest trends in statistical production, continuous use of information acquired in regular CSB surveys and proportionate reduction of the response burden are among the key CSB priorities.

In cooperation with holders of administrative data and in line with the competences provided for in the Statistics Law, CSB is striving to solve the issues related to the use of administrative data sources, thus aiming to acquire as comprehensive and high-quality administrative data allowing to reduce response burden on enterprises and households as possible.

CSB measures to improve use of administrative data and reduce response burden taken in 2020 (in Latvian only).

CSB Revision policy has been lain down in accordance with the European Union and international recommendations and good practice.

Basically, the CSB statistical data are revised in accordance with the scheduled revision cycle and plan.

Within the context of industry statistics routine revisions of statistical data concern:

• regular revisions of the published data, adding revised or more detailed information (revision of data of the previous quarter); revisions of the published data with the help of seasonal adjustment method.
• long-term planned revisions of the published data in accordance with the changes in methodology or classifications.
• unplanned revision (correction of errors) are carried out, if a respondent updates, supplements or corrects data on any of the previous periods, or if the methodology changes.

Vintage database has not introduced.

Revision policy is available on the CSB website: https://www.csp.gov.lv/en/media/5105/download

The same revision policy is applied nationally and in transmissions to Eurostat.

Non- adjusted data (equal to calendar adjusted data) on previous quarter could be revised. When Construction Production Index for a new quarter is published, the data of previous quarter could be updated. In practice data are not revised (MR=0 and MAR=0).

Seasonally adjusted figures are revised from January 2000 onwards.

Mean Revision is defined as average of the difference between the first published data for each quarter and the latest published data for that quarter, calculated for the last 20 quarters.

L_t -Latest estimate

P_t - Preliminary estimate

R_t - Revision

n - number of observations

The data are compiled based on the CSB quarterly questionnaire on construction (1-construction).

Framework source: Statistical Business Register (SUR)

Type of sample: single - stage sample

Sampling coordination: independent sampling

Stratification variables: target population, industry, size class by number of employees

Allocation: optimal proportional

Sampling: A stratified simple random sample

Target population: In 2022, economically active statistical units whose main activity is NACE Section F, excluding industry 4110 and excluding statistical units whose ISK starts with S13.

The sample includes all enterprises with at least 20 employees and NACE Rev. 2. F 4213 with at least 5 employees. Units with fewer employees are surveyed at random using stratified simple random sampling. Statistical units with less than 5 employees (except NACE Rev. 2 F 4322, 4332, 4334, 4339 - with less than 20 employees and NACE Rev. 2 F 4333, 4399 with less than 7 employees) are valued, using administrative data (volume of VAT transactions).

The number of units in each part of the population varies. It is adjusted to include new economically active enterprises (according to the above criteria) in the set of surveyed statistical units, specifying the economic activities of enterprises, receiving additional or clarified information from respondents, as well as more up-to-date data from administrative data sources.

Quarterly

Data on construction output are collected using Form 1-Construction "Report on self-construction by enterprises". Most data is submitted in the form of an e-report, other forms of submission (mail, e-mail, telephone) are less common.

When the form or automatic validation controls change, training for data collectors (statisticians) is provided, as well as instructions are prepared for both respondents and statisticians.

Respondents receive notifications and reminders about the need to submit a report if the submission deadline is approaching. If the respondents do not submit the report in time, they will be contacted by statisticians. Each report (respondent) not submitted on time is coded with an appropriate non-response code with an explanation attached to it.

When connecting to the e-reporting system and selecting the appropriate report, the information about the applicant is filled in automatically and only the information needs to be checked and corrected (in case of changes).

CSB has the main program used for data entry, processing with automatic control and validation tools - ISDMS (Integrated Statistical Data Management System).
Standard MS Office professional software (MS Excel, MS Access) is used daily for the compilation of construction statistics.

Before the data are transmitted to Eurostat the data are validated and modified in the SDMX format.

Index of Production in Construction (henceforth IPC) is calculated using chain-linked method with the annual overlap technique, where calculation is based on the average previous year’s quarterly volume of construction works (at constant prices) and value added accumulated by enterprises two years ago in breakdown by NACE divisions is used as weights. The IPC is calculated by recalculating production output indicator at constant prices with the help of construction cost indices. The weights are changed every year, thus changes in the structure of construction are taken into account. Average volume of output in construction in 2015 is used as the base for the comparison.

Imputation is the process used to assign replacement values for missing, invalid or inconsistent data that have failed edits. This includes automatic and manual imputations; it excludes follow-up with respondents and the corresponding corrections (if applicable). Thus, imputation as defined above occurs after data collection, no matter from which source or mix of sources the data have been obtained, including administrative data. 

This indicator is influenced both by the item non-response and the editing process. It measures both the relative amount of imputed values and the relative influence on the final estimates from the imputation procedures. The unweighted imputation rate for a variable is the ratio of the number of imputed values to the total number of values requested for the variable. The weighted rate shows the relative contribution to a statistic from imputed values; typically a total for a quantitative variable. For a qualitative variable, the relative contribution is based on the number of units with an imputed value for the qualitative item.

The production value data are deflated with the construction cost index of the corresponding NACE Rev.2 class.
Calendar and seasonal adjustment are calculated using TRAMO/SEATS method.

Calendar adjusted data - data in which the influence of varying number of working days, leap year and rolling weekends has been eliminated.

Seasonally adjusted data (construction) - data that in time series analysis allow assessing changes in economic processes more objectively, eliminating influence of seasonality and calendar effect. This method allows reducing, e.g., influence of Easter, Christmas and other holidays, as well as fluctuations in production volume of sectors depending on the season of construction.
It must be taken into account that in accordance with current seasonal adjustment methodology, when data on the new period are added, the data on previous period are recalculated; therefore information on previous periods in the adjusted time series may change each quarter.

Not available.